The Increasing Expression of GPX7 Related to the Malignant Clinical

Yao et al. Chinese Neurosurgical Journal (2021) 7:21 https://doi.org/10.1186/s41016-021-00235-3 中华医学会神经外科学分会 CHINESE MEDICAL ASSOCIATION CHINESE NEUROSURGICAL SOCIETY RESEARCH Open Access The increasing expression of GPX7 related to the malignant clinical features leading to poor prognosis of glioma patients Jiawei Yao1,2†, Xin Chen1,2†, Zhendong Liu1,2†, Ruotian Zhang1,2†, Cheng Zhang5, Quan Yang1,2, Penglei Yao1,2, Qiuyi Jiang1,2, Jianing Wu1,2* and Shiguang Zhao1,2,3,4*

Abstract Background: Glioma is the most common malignant brain tumor in adults. The standard treatment scheme of glioma is surgical resection combined alternative radio- and chemotherapy. However, the outcome of glioma patients was unsatisfied. Here, we aimed to explore the molecular and biological function characteristics of GPX7 in glioma. Methods: The multidimensional data of glioma samples were downloaded from Chinese Glioma Genome Atlas (CGGA). RT-qPCR method was used to identify the expression status of GPX7. Kaplan–Meier curves and Cox regression analysis were used to explore the prognostic value of GPX7. Gene Set Enrichment Analysis (GSEA) was applied to investigate the GPX7-related functions in glioma. Results: The results indicated that the expression of GPX7 in glioma was higher compared to that in normal brain tissue. Univariate and multivariate Cox regression analyses confirmed that the expression value of GPX7 was an independent prognostic factor in glioma. The GSEA analysis showed that GPX7 was significantly enriched in the cell cycle pathway, ECM pathway, focal adhesion pathway, and toll-like receptor pathway. Conclusions: The GPX7 was recommended as an independent risk factor for patients diagnosed with glioma for the first time and GPX7 could be potentially used as the therapy target in future. Furthermore, we attempted to explore a potential biomarker for improving the diagnosis and prognosis of patients with glioma. Keywords: GPX7, Glioma, Biomarker, Survival

* Correspondence: [email protected]; [email protected] †Jiawei Yao, Xin Chen, Zhendong Liu and Ruotian Zhang contributed equally to this work. 1Department of Neurosurgery, First Affiliated Hospital of Harbin Medical University, No. 23 Youzheng Street, Nangang District, Harbin 150001, Heilongjiang Province, China Full list of author information is available at the end of the article

© The Author(s). 2021 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Yao et al. Chinese Neurosurgical Journal (2021) 7:21 Page 2 of 9

Background Here, we investigated the molecular and clinical fea- Glioma is a malignant tumor with the highest morbidity tures of GPX7 in glioma through bioinformatics analysis. and mortality in the central nervous system, which rep- Meanwhile, we attempted to reveal the GPX7-related resented up to 81% of central nervous system malignan- pathological progress in gliomas. At the same time, we cies [1, 2]. The current standard treatment is surgical endeavored to establish a novel therapeutic target to im- resection, combined radiotherapy and chemotherapy [3]. prove the diagnosis and the prognosis of glioma. Although active treatments are taken, the overall survival (OS) and progression-free survival of glioma patients is still unsatisfying and unpredictable [4].. Therefore, it is Methods imperative to identify the novel prognostic biomarkers RNA sequencing and clinic information from CGGA data for glioma. repository Considerable studies have demonstrated that mRNA is We downloaded the clinical and multidimensional data involved in the molecular regulation mechanism of gli- of more than 1000 Chinese glioma patients from the oma. Thus, mRNA can be used as a potential biomarker Chinese Glioma Genome Atlas (CGGA) (http://www. to predict the overall survival time of glioma patients cgga.org.cn/). The data included the whole-exome se- [5]. For instance, LGALS3, L1CAM, and SCAMP3 were quencing (286), DNA methylation (159), mRNA sequen- associated with the shorten OS of glioma patients by cing (1018), mRNA microarray (301), microRNA promoting the proliferation or other malignant tumor microarray (198), and matched clinical data. In CGGA characteristics of glioma [6–8]. Despite previous studies database, the RNA sequencing data and clinical informa- partially revealed that the abnormal expression of tion of 325 and 693 patients were obtained in two ses- mRNA is implicated in the occurrence and development sions. The sva package and limma package in R language of glioma via multiple mechanisms, the pathogenesis of were used to intersect the genes of the two data sets and these mRNAs remains limited due to the complex bio- merge the obtained data into one data set for subsequent logical characteristics in gliomas. Therefore, the aim of analysis. Further, the glioma samples with incomplete this study was to explore a crucial mRNA for glioma to clinical information were eliminated. And finally, 749 predict the prognosis of glioma patients. patients were enrolled in our study. Mammalian GPX7 is a non-selenocysteine containing Gene expression profiling interactive analysis (GEPIA) phospholipid peroxide glutathione peroxidase [9] and database (http://gepia.cancer-pku.cn/) is a web server for plays a significant role in maintaining redox status [10]. gene expression profiling and interaction analysis of can- Therefore, the body avoids entering an oxidative stress cer and normal tissues. A total of 163 samples of glio- state. Numerous studies have revealed that oxidative blastoma, 518 samples of low-grade glioma, and 207 stress state exhibited an inhibitory effect on the patho- normal samples were enrolled in our study. logical process of tumors. For example, miRNA-338-5p promotes oxidative stress by regulating the Hedgehog GSEA pathway and ultimately inhibited the proliferation of gli- Gene Set Enrichment Analysis (GSEA) is a useful tool omas [11]. Moreover, inducing oxidative stress of C6 gli- which could be used to analyze the functions and signal oma cells was related with the promotion of cell cycle pathways of genes. The samples downloaded from the GPX7 arrest and apoptosis [12]. However, whether was CGGA database were separated into high GPX7-expres- involved in the pathological process of tumors remains sion group and low GPX7-expression group. A p value GPX7 unknown. Peng et al. showed that was absent in of less than 0.05 and an FDR value of less than 0.25 were esophageal adenocarcinoma and was associated with significant. methylation of specific sites in the exon region of GPX7. Besides, the biological effect of overexpressing GPX7 on inhibiting cell proliferation and promoting cell senes- RT-qPCR cence has been verified in vitro experiments and in vivo Total RNA was extracted from gliomas and normal con- models [13]. However, the expression of GPX7 was also trol tissues by tri-Regent (Sigma, USA) (samples source: reported as an oncogene. Nonetheless, recent studies the First Affiliated Hospital of Harbin Medical Univer- have showed that overexpression of GPX7 could predict sity), RT-qPCR was performed using fast start universal malignant entities in liver cancer [14]. Hence, the role of SYBR Green Master (Rox) (Roche, Germany), -Δ CT GPX7 in tumors remains controversial. So far, there is reckoned the mRNA expression. The sequence of GAPD no relevant report of the function of GPX7 in glioma. H was 5′-CAAGGTCATCACTGATGAACTTTG-3′ (F) Thus, the role of GPX7 in glioma, whether it could pro- and 5′-GTCCACCCTGTTGCTGTAG-3′ (R), and the mote the pathological process of glioma or serves as a primer sequence of GPX7 was 5′-TCACAGACCACTAC clinical therapeutic target, remained elusive. CGA-3′ (F) and 5′-CGGGGACACTACTCATTC-3′ (R). Yao et al. Chinese Neurosurgical Journal (2021) 7:21 Page 3 of 9

Statistical analysis group and a low GPX7 expression group. The survival The transcriptome data from glioma samples from the package and survminer package were used to analyze CGGA database were comprehensively analyzed. the survival differences between two group. The five- Kaplan–Meier method and Cox regression model were year survival rate of the GPX7 overexpression group was used to exploring the prognostic value of GPX7 in gli- 14.7% (95% CI [0.1107–0.196]) while the 5-year survival oma patients. Gene set enrichment analysis (GSEA) was rate of the GPX7 low expression group was 63.3% (95% used to predict the GPX7-related biological functions CI [0.583–0.689]) (Fig. 1a). and pathways in gliomas. Co-expression analysis was The receiver operating characteristic curve (ROC) used to detect the genes most related to the expression was used to further verify whether the expression of GPX7. level of GPX7 was reliable in predicting the survival time of patients. The AUC was more than 0.7 in the Results 1-year, 3-year, and 5-year ROC curve for the prognos- Clinical characteristics tic model (Fig. 1b). The area under the ROC curve The clinical information of gliomas, including PRS type, (AUC) was more significant than 0.7, and which sug- histology, grade, gender, age, radio status, chemotherapy, gested that the GPX7 expression was reliable in pre- IDH mutation status, and 1p19q codeletion status, were dicting survival time. shown in Table S1. Additionally, according to the GEPIA database, the expression of GPX7 was significantly higher in glioblast- High expression of GPX7 shortens the overall survival oma (GBM) and low-grade gliomas (LGG) than normal time in glioma patients tissues. What is more, high GPX7 expression remains as- To explore how GPX7 affects glioma patients, we sepa- sociated with poor prognosis of glioma patients (p < rated the obtained samples into high GPX7 expression 0.05) (Fig. 1c, d).

Fig. 1 The Kaplan–Meier survival curve and The ROC curve. a Highly expressed GPX7 leads to shorter overall survival time for patients with glioma. b ROC curve shows that using GPX7 as a prognostic marker is valuable. c The expression of GPX7 in GBM and LGG was higher than that in normal tissues in the GEPIA database. d Glioma patients with high GPX7 expression have shorter survival times than those with low expression in the GEPIA database Yao et al. Chinese Neurosurgical Journal (2021) 7:21 Page 4 of 9

Overexpressed GPX7 as an independent risk factor for Potential pathway of GPX7 in regulating malignant predicting the prognosis of glioma patients biological behavior of glioma Furthermore, to verify whether GPX7 can be used as an We used the GSEA method to identify the GPX7-related independent risk factor in predicting the prognosis of biological functions. Notably, GSEA is a ubiquitous bio- glioma patients, the Cox regression model was used to informatics analysis tool. This can help researchers iden- assess GPX7 expression and other clinical features. Uni- tify the enrichment pathway of the target gene and variate and multivariate Cox analysis confirmed that indirectly understand the biological function of the tar- GPX7 was an independent risk factor for glioma pa- get gene. Therefore, GPX7 was divided into a high ex- tients. The remaining clinical features associated with pression group and a low expression group. The results the OS of glioma patients are shown in Fig. 2. indicated that the GPX7 expression was related to multiple tumor-related pathways, including the cell cycle pathway, ECM pathway, focal adhesion pathway, and The relationship between GPX7 and different clinical toll-like receptor pathway (Fig. 4a–d). These pathways characteristics were thus considered as potential strategies for GPX7 to Based on previous studies, the factors that might in- achieve its regulation in glioma. fluence the prognosis of glioma patients were identi- fied. Then, we used the Wilcox.test and Kruskal.test Co-expression analysis to investigate the relationship between GPX7 and To further clarify the role of GPX7 in the pathophysi- clinical factors. The results revealed that the expres- ology of gliomas, we found the GPX7-related genes with sion of GPX7 was positively correlated with PRS type, the Pearson correlation test. Top five genes which were history, grade, age, and chemotherapy, while it was positively correlated with GPX7, including DDOST, negatively correlated with IDH mutation status and RCC2, MCM3, MAGOH, and HDAC1, were selected. 1p19q codeletion status (Fig. 3a-g). Meanwhile, the top five genes which were negatively

Fig. 2 Relationship between clinical features and prognosis of patients with glioma. a Univariate analysis. b Multivariate analysis Yao et al. Chinese Neurosurgical Journal (2021) 7:21 Page 5 of 9

Fig. 3 Relationship between GPX7 and clinical characteristics. The expression of GPX7 is positively correlated with a PRS type, b age, c grade, d chemotherapy, and g histology. The expression of GPX7 is negatively correlated with e IDH mutation and f 1p/19q codeletion correlated with GPX7, including RIMS1, AMER3, overexpression indicated a poor prognosis for gliomas. TMEM56, SVOP, and FBXW4 were selected. As shown For further verification, 5 glioma samples and 5 normal in Figs. 5 and 6, the result suggested that GPX7 might brain tissue samples were used to verify the expression promote the expression of genes that are positively cor- level of GPX7 by RT-qPCR. As a result, the expression related with its expression level or that these co- of GPX7 in gliomas was higher compared to that in nor- expressed positively related genes co-promote or inhibit mal tissues (p < 0.05) (Fig. 7), which corroborated with specific pathophysiological processes with GPX7 to pro- the findings from previous analysis of the CGGA mote the occurrence and development of glioma. Con- database. versely, GPX7 might inhibit the expression of genes that are negatively correlated, or the function between them Discussion is antagonistic. Reports indicate that GPX7 functions distinctly in different tumors. For instance, it is a tumor suppressor gene Validating GPX7 expression level in glioma using RT-qPCR in esophageal adenocarcinoma but a carcinogenic gene By analyzing the data from the CGGA database, it was in liver cancer [13, 14]. Nonetheless, the role of GPX7 in concluded that GPX7 is a novel oncogene, and its glioma remains unknown. As such, to reveal the role of Yao et al. Chinese Neurosurgical Journal (2021) 7:21 Page 6 of 9

Fig. 4 Significantly enriched pathways with GSEA. a Cell cycle pathway. b ECM pathway. c Focal adhesion pathway. d Toll-like receptor pathway

Fig. 5 Gene expression heat map and correlations for GPX7 co-expressed genes Yao et al. Chinese Neurosurgical Journal (2021) 7:21 Page 7 of 9

Fig. 6 The co-expressed network of GPX7. a The circos plot was constructed by 10 genes most related to GPX7 expression, including 5 positively related genes and 5 negatively related genes (p < 0.05). b Co-expression relationship of 10 genes in GPX7

GPX7 in glioma, this work first obtained RNA sequen- GPX7 could still be used as an independent risk factor cing data of more than one thousand glioma patients in predicting the prognosis of glioma patients after ex- from the CGGA database. cluding the influence of other factors. At the same time, Through the correlation analysis between GPX7 and our analysis demonstrated that other clinical characters, OS of glioma patients, we discovered that the is the ex- including PRS type, grade, age, chemotherapy, IDH mu- pression of GPX7 was negatively related with the OS in tation status, and 1p19q codeletion status, can also act glioma patients. It is of pivotal importance to verify the as independent risk factors. Moreover, we uncovered reliability of GPX7 in predicting the OS of glioma pa- that the expression level of GPX7 is related to different tients by the ROC curve. The overexpression of GPX7 in clinical characteristics, thereby indicating that overex- gliomas was further validated by the GEPIA database pression of GPX7 might influence the malignant pro- and it was associated with poor prognosis. Nevertheless, gress of these clinical characteristics via some potential the occurrence and development of glioma are induced mechanisms. Based on the above studies, we proved for by the imbalance between multiple carcinogenic and the first time that GPX7 could be used as a novel onco- cancer suppressive factors; hence, several factors are im- gene in predicting the prognosis of glioma patients. plicated in this complex pathological process. Whether Nonetheless, the mechanism of GPX7 in regulating the GPX7 can be used as an independent risk factor for gli- malignant behavior of glioma remains controversial. oma patients remains unclear. GSEA analysis suggested that GPX7 might involve in Univariate and multivariate analyses were performed joining the pathophysiological process of glioma. The by the Cox regression model. As a result, we found that outcomes revealed some significantly enriched pathways, including cell cycle pathway, ECM pathway, focal adhesion pathway, and toll-like receptor pathway, Catchin flavonoids inhibited glioma by blocking cell cycle signaling pathways [15]. The invasiveness of glioma cells can be achieved by selectively interacting with the extracellu- lar matrix (ECM) components [16]. Mda-9/syntenin promotes glioma migration through focal adhesion kinase (FAK) in the focal adhesion signaling pathway [17]. The toll-like receptor family is closely related to cancer. Toll- like receptor 4 promotes glioma proliferation by activat- ing the NF-kappa B signaling pathway [18]. Toll-like receptor 2 regulates glioma invasion by promoting the expression of matrix metalloproteinases [19]. The above pathways promote the pathological process of glioma through proliferation, migration, invasion, and other malignant biological behaviors of glioma cells. The enrichment of GPX7 in these pathways revealed that glioma Fig. 7 Detection of GPX7 expression in gliomas by RT-qPCR might participate in the regulation via various complex Yao et al. Chinese Neurosurgical Journal (2021) 7:21 Page 8 of 9

pathways. Despite reaching a preliminary conclusion on Supplementary Information how GPX7 regulates glioma by GSEA analysis, the bio- The online version contains supplementary material available at https://doi. org/10.1186/s41016-021-00235-3. logical behavior of glioma cannot be fully regulated by a single factor due to the complex pathological process of Additional file 1: Table S1. Characteristics of patients with glioma glioma. based on CGGA. Description of data: Clinical information of the samples Therefore, we screened 10 genes having the highest we used for analysis. correlation with GPX7 expression by co-expression analysis, among which 5 genes had a positive correlation, in- Abbreviations cluding DDOST, RCC2, MCM3, MAGOH, and HDAC1. CGGA: The Chinese Glioma Genome Atlas; GEPIA: Gene Expression Profiling Interactive Analysis; OS: Overall survival; GSEA: Gene Set Enrichment Analysis; The pathophysiological processes regulated by these ROC: Receiver operating characteristic curve; AUC: Area under the ROC curve genes were closely related to tumors. The high expression of DDOST in colon adenocarcinoma promotes the Acknowledgements Not applicable. proliferation of tumor cells [20]. The expression of RCC2 in ER-positive breast induces the expression of Authors’ contributions IGF-1 and leads to the malignant behavior of tumor JW Y and XC designed and guided the subject. ZD L and RT Z collated and cells, including proliferation, invasion, etc. [21]. MCM3 analyzed the data. CZ and QY completed the writing and repair of the manuscript. PL Y and QY J completed the validation experiment part. All is an independent risk factor for invasive ductal carcin- authors reviewed and approved the final manuscript. oma [22]. MAGOH regulates melanoma production by controlling the proliferation of melanocyte [23]. HDAC1 Funding promotes liver cancer metastasis via the FAM99A- National Natural Science Foundation of China (81972363). miR92a signaling pathway [24]. Additionally, the 5 nega- Availability of data and materials tively related genes include RIMS1, AMER3, TMEM56, All data generated or analyzed during this study are included in this SVOP, and FBXW4. Therefore, the occurrence and de- published article. velopment of glioma is a significantly complex biological Ethics approval and consent to participate process, and GPX7 interacts with various oncogenes and Informed consent was obtained from patients to collect tissue samples for tumor suppressor genes to participate in the network of the study, and the study was reviewed and approved by the Ethics GPX7 Committee of the First Affiliated Hospital of Harbin Medical University (No. glioma regulation. is considered as a novel mo- HMUIRB-2008-06). lecular marker that predicts the prognosis of glioma patients, with high validity and reliability. Consent for publication For the first time, through analysis of multidimen- Not applicable. GPX7 sional data, we propose that can be used as a Competing interests novel oncogene to shorten the prognosis of glioma pa- The authors declare that they have no competing interests. tients. Nonetheless, this study has some shortcomings. Author details Firstly, the CGGA database only enrolled more than 1Department of Neurosurgery, First Affiliated Hospital of Harbin Medical 1000 high-throughput data of glioma patients without University, No. 23 Youzheng Street, Nangang District, Harbin 150001, non-glioma individuals. Therefore, all the findings were Heilongjiang Province, China. 2Key Colleges and Universities Laboratory of Neurosurgery in Heilongjiang Province, Harbin 150001, Heilongjiang based on the comparison between the high- and low- Province, China. 3Department of Neurosurgery, The Pinghu Hospital of expression of the GPX7 group. However, we found that Shenzhen University, Shenzhen 518100, Guangdong Province, China. the high expression of GPX7 accounts for malignant 4Institute of Neuroscience, Sino-Russian Medical Research Center, Harbin Medical University, Harbin 150001, Heilongjiang Province, China. 5North progression of glioma by comparing the overall survival Broward Preparatory School, 7600 Lyons Rd, Coconut Creek, FL 33073, USA. time between the two groups and verifying its credibility via ROC curves. The expression of GPX7 was validated Received: 18 August 2020 Accepted: 26 January 2021 in glioma and normal brain tissues by RT-qPCR to com- pensate for the insufficiency of sample information in References the database. Consequently, we found that GPX7 was 1. Zhang C, Cheng W, Ren X, et al. Tumor Purity as an Underlying Key Factor highly expressed in glioma. Notably, this was in line with in Glioma. Clin Cancer Res. 2017;23(20):6279–91. 2. Adel Fahmideh M, Schwartzbaum J, Frumento P, et al. Association between our previous findings. DNA repair gene polymorphisms and risk of glioma: a systematic review and meta-analysis. Neuro Oncol. 2014;16(6):807–14. Conclusion 3. Cheng M, Zhang ZW, JI XH, et al. Super-enhancers: A new frontier for GPX7 glioma treatment. Biochim Biophys Acta Rev cancer. 2020;1873(2):188353. For the first time, the was reported as a carcino- 4. Chen W, Lei C, Liu P, et al. Progress and Prospects of Recurrent Glioma: A genic gene in glioma, and our results revealed the correl- Recent Scientometric Analysis of the Web of Science in 2019. World ation between gliomas and malignant clinical features. Neurosurg. 2020;134:e387–e99. 5. Öberg K, Califano A, Strosberg JR, et al. A meta-analysis of the accuracy of a Moreover, our findings provided a new diagnostic and neuroendocrine tumor mRNA genomic biomarker (NETest) in blood. Ann prognostic marker for study in future. Oncol. 2020;31(2):202–12. Yao et al. Chinese Neurosurgical Journal (2021) 7:21 Page 9 of 9

6. Hu WM, Yang YZ, Zhang TZ, et al. LGALS3 Is a Poor Prognostic Factor in Diffusely Infiltrating Gliomas and Is Closely Correlated With CD163+ Tumor- Associated Macrophages. Front Med. 2020;7:182. 7. Zeng J, Xi SY, Wang F, et al. L1CAM High Expression Associates with Poor Prognosis in Glioma but Does Not Correlate with C11orf95-RELA Fusion. BioMed Res Int. 2020;2020:1353284. 8. Li C, Zhang Z, LV P, et al. SCAMP3 Promotes Glioma Proliferation and Indicates Unfavorable Prognosis via Multiple Pathways. Onco Targets Ther. 2020;13:3677–87. 9. Utomo A, Jiang X, Furuta S, et al. Identification of a novel putative non- selenocysteine containing phospholipid hydroperoxide glutathione peroxidase (NPGPx) essential for alleviating oxidative stress generated from polyunsaturated fatty acids in breast cancer cells. J Biol Chem. 2004;279(42): 43522–9. 10. Chang YC, Yu YH, Shew JY, et al. Deficiency of NPGPx, an oxidative stress sensor, leads to obesity in mice and human. EMBO Mol Med. 2013;5(8): 1165–79. 11. Zhou H, Han L, Wang H, et al. Chidamide Inhibits Glioma Cells by Increasing Oxidative Stress via the miRNA-338-5p Regulation of Hedgehog Signaling. Oxid Med Cell Longev. 2020;2020:7126976. 12. Shukla A, Trivedi SP. An in vitro analysis of the rat C6 glioma cells to elucidate the linear alkylbenzene sulfonate induced oxidative stress and consequent G2/M phase cell cycle arrest and cellular apoptosis. Chemosphere. 2018;205:443–51. 13. Peng D, Hu T, Soutto M, et al. Glutathione peroxidase 7 has potential tumour suppressor functions that are silenced by location-specific methylation in oesophageal adenocarcinoma. Gut. 2014;63(4):540–51. 14. Guerriero E, Capone F, Accardo M, et al. GPX4 and GPX7 over-expression in human hepatocellular carcinoma tissues. Eur J Histochem. 2015;59(4):2540. 15. Zhan Sheng Z. Anticancer effects of catechin flavonoid in human glioma cells are mediated via autophagy induction, cell cycle arrest, inhibition of cell migration and invasion and targeting MAPK/ERK signalling pathway. J BUON. 2020;25(2):1084–90. 16. Goldbrunner RH, Bernstein JJ, Tonn JC. ECM-mediated glioma cell invasion. Microsc Res Tech. 1998;43(3):250–7. 17. Zhong D, Ran JH, Tang WY, et al. Mda-9/syntenin promotes human brain glioma migration through focal adhesion kinase (FAK)-JNK and FAK-AKT signaling. Asian Pac J Cancer Prev. 2012;13(6):2897–901. 18. Jiang Y, Zhou J, Luo P, et al. Prosaposin promotes the proliferation and tumorigenesis of glioma through toll-like receptor 4 (TLR4)-mediated NF-κB signaling pathway. EBioMedicine. 2018;37:78–90. 19. Wang F, Zhang P, Yang L, et al. Activation of toll-like receptor 2 promotes invasion by upregulating MMPs in glioma stem cells. Am J Transl Res. 2015; 7(3):607–15. 20. Hu M, Fu X, Si Z, et al. Identification of Differently Expressed Genes Associated With Prognosis and Growth in Colon Adenocarcinoma Based on Integrated Bioinformatics Analysis. Front Genet. 2019;10:1245. 21. Wang W, Xu B, Zhang Z, et al. RCC2 Expression Stimulates ER-Positive Breast Tumorigenesis. J Oncol. 2020;2020:5619462. 22. Zhao Y, Wang Y, Zhu F, et al. Gene expression profiling revealed MCM3 to be a better marker than Ki67 in prognosis of invasive ductal breast carcinoma patients. Clin Exp Med. 2020;20(2):249–59. 23. Silver DL, Leeds KE, Hwang HW, et al. The EJC component Magoh regulates proliferation and expansion of neural crest-derived melanocytes. Dev Biol. 2013;375(2):172–81. 24. Zhao B, Ke k, Wang Y, et al. HIF-1α and HDAC1 mediated regulation of FAM99A-miR92a signaling contributes to hypoxia induced HCC metastasis. Signal Transduct Target Ther. 2020;5(1):118.